Nonstationary Hazard Modelling

Data Challenge

Zak Varty

Data Challenge

Purpose

  • Put what you have been learning into action
  • Work with new people (3 groups)

Task

  • Explore a dataset relating to an environmental hazard (eathquakes)
  • Work towards answering some meaningful questions.

Deliverable

  • 5-10 minute presentation (email to z.varty@imperial.ac.uk before 14:30.)

About the Data

  • Earthquakes in the NE of Netherlands
  • Caused by extraction of Natural Gas
  • Small but close to the surface

A look at the Data

date time location_X location_Y lat lon mag place type
2024-05-18 01:52:34 245053.1 598471.3 53.365 6.740 0.8 Garsthuizen induced earthquake
2024-05-13 23:11:10 242612.0 582619.1 53.223 6.699 0.4 Harkstede induced earthquake
2024-05-12 11:38:16 254492.1 586412.4 53.255 6.878 0.8 Siddeburen induced earthquake
2024-05-05 19:14:45 247932.5 597635.7 53.357 6.783 0.3 ’t Zandt induced earthquake
2024-04-14 14:08:51 242502.6 584954.7 53.244 6.698 1.6 Lageland induced earthquake
2024-04-11 09:10:24 245078.2 597135.9 53.353 6.740 2.1 Zeerijp induced earthquake

Spatial Snapshot

Temporal Snapshot

Questions

What statistical models would be suitable for modelling the rate of large earthquakes over time / space / space-time?

  1. Is the rate of (large) earthquakes changing over time?
  2. Are the magnitudes of earthquakes changing over time?
  3. Is the distribution of large earthquake magnitudes consistent with the Gutenberg-Richter law (exponential distribution)?
  4. Is the rate / size of earthquakes the same across space?
  5. Can we define something like a return-level to quantify seismic hazard?
  6. Can we produce a map of seismic hazard over space / time?

Tips

  • Aim for 3 - 5 slides.

  • Don’t worry about polishing graphs. Correct >> Pretty.

  • 1 - 2 people to present.

  • Tell us about:

    • what you found,
    • what went well,
    • what could be better.

Build Information

R version 4.3.3 (2024-02-29)

Platform: x86_64-apple-darwin20 (64-bit)

locale: en_US.UTF-8||en_US.UTF-8||en_US.UTF-8||C||en_US.UTF-8||en_US.UTF-8

attached base packages: stats, graphics, grDevices, utils, datasets, methods and base

other attached packages: multipanelfigure(v.2.1.5), ggtext(v.0.1.2), colorspace(v.2.1-0), dplyr(v.1.1.4) and ggplot2(v.3.5.0)

loaded via a namespace (and not attached): utf8(v.1.2.4), generics(v.0.1.3), xml2(v.1.3.6), stringi(v.1.8.3), hms(v.1.1.3), digest(v.0.6.35), magrittr(v.2.0.3), evaluate(v.0.23), grid(v.4.3.3), timechange(v.0.3.0), showtextdb(v.3.0), sysfonts(v.0.8.9), fastmap(v.1.1.1), jsonlite(v.1.8.8), assertive.files(v.0.0-2), pander(v.0.6.5), fansi(v.1.0.6), scales(v.1.3.0), cli(v.3.6.2), zvplot(v.0.0.0.9000), rlang(v.1.1.3), crayon(v.1.5.2), commonmark(v.1.9.1), bit64(v.4.0.5), munsell(v.0.5.1), withr(v.3.0.0), yaml(v.2.3.8), tools(v.4.3.3), parallel(v.4.3.3), tzdb(v.0.4.0), showtext(v.0.9-7), vctrs(v.0.6.5), R6(v.2.5.1), gridGraphics(v.0.5-1), lifecycle(v.1.0.4), lubridate(v.1.9.3), magick(v.2.8.3), bit(v.4.0.5), vroom(v.1.6.5), assertive.base(v.0.0-9), pkgconfig(v.2.0.3), pillar(v.1.9.0), gtable(v.0.3.4), glue(v.1.7.0), Rcpp(v.1.0.12), xfun(v.0.43), tibble(v.3.2.1), tidyselect(v.1.2.1), rstudioapi(v.0.16.0), knitr(v.1.45), farver(v.2.1.1), htmltools(v.0.5.8.1), assertive.numbers(v.0.0-2), labeling(v.0.4.3), rmarkdown(v.2.26), readr(v.2.1.5), compiler(v.4.3.3), markdown(v.1.12) and gridtext(v.0.1.5)

Nonstationary Hazard Modelling - Data Challenge - Zak Varty